Based on the immunobiology of tumor progression in melanoma, the goal of this project is to develop a strategy of "therapeutic vaccination" using genetically engineered melanoma cells for the treatment of patients with disseminated disease. Relevant aspects of the biology and therapeutic implications of melanoma cells as antigen presenting cells (APCs) will be explored in three specific aims. The first aim is to test the hypothesis that the cells of melanocytic neoplasia express costimulatory molecules of the B7 family in a stage-specific fashion; paralleling the waxing and waning T cell filtration of lesions, expression of B7 molecules appears in early primary melanomas and certain precursor lesions and disappears in advanced disease. Biopsies of patients' lesions and cultured melanocytic cells will be examined by mRNA in situ hybridization an by immunohistochemistry for the presence of costimulatory molecules. Concomitantly, expression of MHC class I and II molecules will be measured and the T cell infiltrate characterized. The second aim is to test the hypothesis that melanoma cells of metastatic disease can be enabled as competent APCs in vitro. Initial studies of cell lines and subsequent studies of recently harvested tumor tissue will focus on tumor cells provided with costimulation, first by monoclonal antibodies (Mabs) to the T cell counter-receptor for B7 and then by gene transfer to express B-7-1 (CD80). These cells will be examined in co-cultures with autologous T cells for their capacity to stimulate T cell responses (activation, proliferation, and cytokine generation). We will also examine whether the addition to co-cultures of the multifunctional cytokine, IL012, will foster the development of TH1 cells, the subset of T helper cells that supports the expansion and function of cytolytic T cells (CTL). The third aim is to develop retroviral and adenoviral vectors to mediate the efficient expression of B7-1 by melanoma cells. The extent and duration of expression of B7-1 will be compared in tumor cells transduced with each of these recombinant viral vectors. Animal models will test the feasibility and efficiency of in vivo gene transfer. After direct inoculation of viral vectors into human melanoma xenografts in the skin of immunodeficient mice, the distribution, degree, and persistence of B7-1 expression will be characterized. A second model will examine B7-1 expression after intravenous administration of B7-1 vectors in animals with pulmonary metastases of human melanoma. Vectors for IL-12 expression will be developed an tested in the same fashion. These studies are preliminary to a series of clinical trials testing the safety and efficacy of vaccines comprised of autologous melanoma cells engineered to express costimulation with or without IL-12. The anticipated outcome is tumor destruction mediated by multiple varieties of immune effectors, including but not restricted to TH cells and CTL recognizing the spectrum of antigens expressed on an individual's tumor cells. Definition of the antigens which trigger and are the target of this response, while interesting and important, is not a requirement of this strategy.